RE: Low cost Servo Controller

In response to the questions for more clarification about what a low cost servo controller may offer:

The primary objectives include Cost effective and runs on Windows, DOS, or NT without excessive configuration problems. Linux seems a bit difficult to polish up. On the cost effective side, even in a production shop, if it were not absolutely necessary to install ball screws, one might be able to justify upgrading that dusty knee operated shelf in the corner. No one would ever think of putting balls screws on many of the older mills. If the shop were near a gorge, and the fork lift were handy, a lot of people might walk away with that Calvin and Hobbes grin after dropping these old mills in the gorge. The fact that we can't find some better use for them, than a 2000 pound, adjustable shelf, seems pretty insane if you ask me.

The concept of this back-lash compensation servo driver black box, is that if the software signals are straight foward enough, it may be as simple as a parallel printer port cabel to the computer (on the computer side) and DC brush motors, limit and emergency stop switches and encoders on the table side; and the DC power supply connections (+, gnd, -) via a terminal block.

It is hopeful that the data format this black box gets from the computer is simple and straight foward to minimize the compatability problems with software. It really doesn't matter what format the data is in, if the software and hardware will plug and play.

The stepper control signals came to mind because they are likely easy to understand in terms of minimizing the computer to black box interfacing issues. In servo systems, the feedback likely goes all the way back to program drivers themselves. No doubt servo controllers and boards are expensive and probably have specialized communication requirments which may vary from one to the next. The black box can be the place where the cooridinated and controlled motion takes place without any feedback to the program in the computer. In this manner, the complexity of bringing up the system may be reduced, however, there are a few issues to make that possible.

If the computer is spitting out step and direction from the computer to the motor controller and has no ability to receive any form of status back from the controller, then the black box (and many stepper systems anyway) suffer a disavantage in that whatever goes wrong can not profide feedback to the program itself. The potential advantage of the black box is that it can correct many problems on its own without interacting with the program.

Bertho, you described another potential version of this concept, in that it is possible to design a stepper controller version of this black box, that may use encoders to compensate for missed steps. Dan made a point about this as well. Dan also described the fact that not using ball screws is like manual machining and I agree. I've never intended this to be any more than a small step at a small cost, just above manual machining. I think there are a number of older mills in the hobby shops and in production shops as well that have one last gasp before their tables fall off of the ways and land fills get compacted by their weight.

My brother has an older CNC mill that gets the feedback from the motors themselves and not from the X, Y, and Z movements. He writes his programs to deal with the backlash, and yet he isn't willing to put ball screws in it (the real answer to the problem), because he can't justify the cost of the ball screws. When this concept is used without ball screws, programs must be written as if one were manually milling. The black box between the computer and the older mill will have to be smart enough to solve the problems it is up against in this configuration if the goal of bringing older 2000 pound shelves into lite duty production uses, is to be realized.

The question is, would anyone want to retrofit older mills for lite duty production, if the total cost was less than the cost of the ball screws alone? You see if one is going to put encoders on the tables, then one might as well drive the tables with DC servo motors anyway. One may use ball screws where it is justified with the same controller.

David Howland



-----Original Message-----
From: Bertho Boman [SMTP:boman@...]
Sent: Monday, September 13, 1999 3:26 PM
To: CAD_CAM_EDM_DRO@onelist.com
Subject: Re: [CAD_CAM_EDM_DRO] Low cost Servo Controller

From: Bertho Boman <boman@...>

David and the gang,

I am not sure what the real goal is that you are proposing. I see several possibilities:

1. If a real servo amp is available but the available software is only written for steppers. The stepper count would internally in a "black box" be compared to the encoder positions and a feed back sent to the servo amp.

2. A servo amp is to be designed from scratch to use stepper control signals.

3. If the stepper software is smart enough to use encoders for feed back and slip compensation, there is a third possibility: A "black box" could be build that converts stepper pulses to +/- 10 VDC to control a standard servo amp.

This brings us to question #1: Is this done just to be able to run servos on stepper software or is an attempt to get a lower cost servo system?

I like to aim for a real servo system using real servo software. The first software that comes to my mind for "home" use is EMC. What other ones are available for let's say below $1,000?

General Comments:

I am sure most of you know this but there seems to be questions raised so here it is:

Often it is claimed that a "servo motor" must be used. A plain DC motor is fine most of the time since we are not trying to start/stop on a dime, running a tape drive or something low inertia, where the motor inertia becomes significant. A "servo motor" often has a built in shaft encoder but that is not required. An encoder is required somewhere in the system but it can be linear encoders, shaft encoders attached in one way or other to the lead screw. The same is true for the tachometer for velocity feed back. We are interested in machining so mounting a tach or encoder to a shaft should not be a big problem.


Low cost servo functional breakdown:

1: Software- EMC is free.

2: I-O. Servo-to-Go has a card with both analog output and shaft encoder inputs. It does all the functions needed but is reasonably priced for industrial use but a little expensive for home use. There are also several commercial low cost boards available that have multiple analog I-O and other boards with just digital I-O. Two separate boards, @ $150 each or so, is still a lot cheaper than the Servo-to-Go. I have not looked at the details of the STG board and I am not trying to be negative about them. I am even likely to buy one shortly from them.

There seems to great opportunities for doing a "joint" low cost I-O board, either through the parallel port or as a plug in card.

3. Servo Amp this is the difficult part. The design and construction is rather complex and not easy as a home project. Jon Elson did a home brew one and he can comment on it better than I can. There is not a lot of them available on the surplus market either.

4. Encoders: Lots of surplus and options. By the way, a "real" stepper system should have encoders too. Other wise it can not verify and guarantee position.

5. Tach. A little harder to find but they are available on the surplus market. The other option is to use a software tach from the encoder pulses.

5. Motors: If we stick with brush motors, there are tons of them available basically for free. The brushless ones are newer, harder to find and trickier to build servo amps for. We are not likely to wear out too many brushes on hobby projects.

As you see, most items are available for peanuts and easy to obtain except the servo amps which are the real difficult ones. The I-O also needs a low cost solution but that is easy.

Project Goals:
Even if we are not producing hardware yet, let's hear your comments and suggestion.

1. Suggestions for real servo software? What are the choices? Obviously EMC is first on the list although I wish I could run it under NT (If you have not noticed, half the messages are Linux help ........ not EMC help) No flames please.

2. I-O board options. STG is one. "Computer Boards" have a lot of I-O boards that might work. What else is available commercially? Anyone already working on one? Who can help?

3: Servo amp. What is available commercially? I bought a few some years ago but the info is out of date. Anyone already working on one? Who can help? Please read my earlier post if you missed it about DSP-PWM servo amps.

If we solve these three questions, we will have a "standard" very high performance system without having to try to create a "hybrid" of something in between.

I am willing to help moving this forward so let us all get together and get it done.

Bertho Boman
==============================================

David Howland wrote:

> From: David Howland <dhowland@...>
>
> Peter, Bertho, Andy, Steve, John, Kirk, Jon thanks for your help. I'm tracking down what I can. Do you have a pointer to the Servo to go card?
>
> Some thoughts I had over the weekend include:
> --------> Snip
>
> So the concept is forming here of a DC Servo Controller that allows feedback from the tables or the motors and responds to Step and Direction signals as if it were a Stepper system. The two advantages are that it is faster and that it is possible to close the loop to the table mounted position data improving the accuracy by overcomming the backlash in the system. Does that concept make since?
>
> Meanwhile I'm looking into the existance of this sort of thing so that I won't have to design one. Please keep the dialog going. Thanks.
>
> David Howland

--------------------------- ONElist Sponsor ----------------------------

ATTENTION ONElist MEMBERS: Get your ONElist news!
Join our MEMBER NEWSLETTER here:
<a href=" http://clickme.onelist.com/ad/newsletter2 ">Click Here</a>

------------------------------------------------------------------------
Welcome to CAD_CAM_EDM_DRO@...,an unmoderated list for the discussion of shop built systems in the above catagories.
To Unsubscribe, read archives, change to or from digest.
Go to: http://www.onelist.com/isregistered.cgi
Log on, and you will go to Member Center, and you can make changes there.
For the FAQ, go to http://www.ktmarketing.com/faq.html
bill,
List Manager